Omnigraph: Rich Representation and Graph Kernel Learning

11 Pages Posted: 2 Mar 2020

See all articles by Boyi Xie

Boyi Xie

Columbia University

Rebecca Passonneau

Center for Computational Learning Systems

Date Written: October 10, 2015

Abstract

OmniGraph, a novel representation to support a range of NLP classification tasks, integrates lexical items, syntactic dependencies and frame semantic parses into graphs. Feature engineering is folded into the learning through convolution graph kernel learning to explore different extents of the graph. A high-dimensional space of features includes individual nodes to complex networks. In experiments on a text-forecasting problem that predicts stock price change from news for company mentions, OmniGraph beats several benchmarks based on bag-of-words, syntactic dependencies, and semantic trees. The highly expressive features OmniGraph discovers provide insights into the semantics across distinct market sectors. To demonstrate the method’s generality, we also report its high performance results on a fine-grained sentiment corpus.

Keywords: machine learning, natural language processing, sentiment analysis, artificial intelligence, financial forecasting, asset pricing, computational finance, quantitative finance, stock price prediction, equity market, financial news

JEL Classification: C53, C55, C63, G11, G12, G14

Suggested Citation

Xie, Boyi and Passonneau, Rebecca, Omnigraph: Rich Representation and Graph Kernel Learning (October 10, 2015). Available at SSRN: https://ssrn.com/abstract=3530575 or http://dx.doi.org/10.2139/ssrn.3530575

Boyi Xie (Contact Author)

Columbia University ( email )

3022 Broadway
New York, NY 10027
United States

Rebecca Passonneau

Center for Computational Learning Systems ( email )

New York, NY
United States

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
9
Abstract Views
122
PlumX Metrics